University of Connecticut OpenCommons@UConn Master's Theses University of Connecticut Graduate School 8-12-2014 Conservation of Bridle Shiner (Notropis bifrenatus) in Connecticut: Issues in Detecting an Elusive Species Kasey C. Pregler [email protected] Recommended Citation Pregler, Kasey C., "Conservation of Bridle Shiner (Notropis bifrenatus) in Connecticut: Issues in Detecting an Elusive Species" (2014). Master's Theses. 677. https://opencommons.uconn.edu/gs_theses/677 This work is brought to you for free and open access by the University of Connecticut Graduate School at OpenCommons@UConn. It has been accepted for inclusion in Master's Theses by an authorized administrator of OpenCommons@UConn. For more information, please contact [email protected]. Conservation of Bridle Shiner (Notropis bifrenatus) in Connecticut: Issues in Detecting an Elusive Species Kasey C. Pregler B.S., University of Connecticut, 2007 A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science at the University of Connecticut 2014 i APPROVAL PAGE Master of Science Thesis Conservation of Bridle Shiner (Notropis bifrenatus) in Connecticut: Issues in Detecting an Elusive Species Presented by Kasey C. Pregler, B.S. ii Acknowledgements Major funding and personnel that facilitated this research were provided by the Connecticut Income Tax Check-off Grant, and the Inland Fisheries Division of the Connecticut Department of Energy & Environmental Protection with additional support from the University of Connecticut through the Storrs Agricultural Experiment Station and Departments of Natural Resources and the Environment, and Ecology and Evolutionary Biology. I acknowledge assistance with field data collection from several seasonal Inland Fisheries Division employees; Matt Traceski, Megan Cruz, Kevin Job, Katie Abbott, Corey Pelletier, Matt Hecklinger, Matt Smith, Chris Finch, Heidi Fitzpatrick, Joe Geenty, and Eric Lindquist. As well as additional assistance from Jan-Michael Hessenauer, Christopher Colgan, Mark Leonard, and Jonathan Velotta who volunteered with seine sampling surveys. Access to survey sites would not have been possible without the willingness and cooperation of the Avalonia Land Conservancy, the town of North Stonington, and many private landowners. This project could not have occurred without the groundwork research from Tim Jensen who assisted with field help and modeling work. I am also thankful for support from my committee members, Eric Schultz, Jason Vokoun, and Chris Elphick and honorary committee member Neal Hagstrom. iii TABLE OF CONTENTS CHAPTER 1: Using multi-method occupancy estimation models to quantify gear differences in detection probabilities: is electrofishing missing occurrences for a species of concern?....................................................................................................................................1 Abstract…………………………………………………………………………………………………….1 Introduction……………………………………………………………………………………………...2 Methods……………………………………………………………………………………………………6 Results……………………………………………………………………………………………………10 Discussion……………………………………………………………………………………………….11 Literature Cited………………………………………………………………………………………17 Tables and Figures………………………………………………………………………………….23 CHAPTER 2: Landscape correlates of localized extirpation in Bridle Shiner (Notropis bifrenatus)…………………………………………………………………………………………………………33 Abstract…………………………………………………………………………………………………..33 Introduction……………………………………………………………………………………………34 Methods………………………………………………………………………………………………….37 Results……………………………………………………………………………………………………41 Discussion……………………………………………………………………………………………….42 Literature Cited……………………………………………………………………………………….46 Tables and Figures…………………………………………………………………………………..52 iv CHAPTER 1 Using multi-method occupancy estimation models to quantify gear differences in detection probabilities: is electrofishing missing occurrences for a species of concern? ABSTRACT Bridle Shiner (Notropis bifrenatus) is a small, rare minnow species native to Northeastern streams and lakes. It is declining over most of its range, and currently is listed as a species of concern in Connecticut. Surveys conducted by the Connecticut state environment agency in the 1960s found Bridle Shiner at 56 locations statewide using seine nets. In contrast, surveys conducted by the agency in the 1990s, using backpack electrofishing, detected Bridle Shiner at only 8 locations. Different sampling techniques made it difficult to assess what portion of the observed decline might be a sampling artifact, confounding efforts to assess the actual conservation status. I sampled 18 habitat patches in three Connecticut watersheds in 2012 to determine if seining for Bridle Shiner yielded a higher detection probability than electrofishing. A multi-method occupancy estimation modeling approach, using program PRESENCE, quantified the probability of correctly detecting Bridle Shiner by gear, and as it covaried with habitat features. Electrofishing detection probability was lower and approximately half that of seining. The abundance of Bridle Shiner in the patch was the most supported covariate to detection, and particularly aided detection for electrofishing. Higher mean water velocity improved detection probability of electrofishing and reduced 1 that of seining. It is possible that the 1990s sampling underestimated the number of populations of Bridle Shiner, and a repeat survey of all historic locations using a seine is recommended. INTRODUCTION Evaluation of sampling methods and gears is an ongoing, if not fundamental endeavor in fisheries conservation and management (Bonar et al. 2009). There exists (and likely always will) a need to find the comparatively most precise and least biased options for quantifying the distribution and abundance of fish species and assemblages (Price and Peterson 2010), and recent efforts towards describing and sharing standard methods among fisheries biologists have emphasized the need for comparability and repeatability among sampling efforts through time and across locations (Bonar et al. 2009). However this can become complicated when uncommon habitats require non-standard methods or use of gear, and when making comparisons to historical data collected with different gears than those currently favored (Patton et al. 1998). Two of the more regularly used gears for freshwater stream fish surveys are seines and electrofishing. Currently, electrofishing is more commonly used than seining and regarded as the most effective gear type when monitoring fish assemblages (Barbour et al. 1998; Poos et al. 2007). Furthermore electrofishing is often the only gear used in wadeable stream-based population assessments and bio- monitoring programs conducted by fisheries management agencies in the United States of America (Kanno et al. 2009). A number of studies have compared seining to backpack electrofishing (Onorato et al. 1998; Poos et al. 2007; Mercado-Silva and 2 Escandón-Sandoval 2013), and all found the backpack to be preferable. However, in shoreline surveys of larger lotic and embayment habitats where boat electrofishing is favored seines have been recommended (Lapointe et al. 2006; Jurajda et al. 2009). The need for adequate sampling methods is especially important for rare, and often imperiled species where assessments can be based on determining the number of occurrences within a political jurisdiction. These are then used as criteria for designating conservation status such as threatened, endangered, etc. (e.g., Master 1991; CT DEEP 2010). Occurrences are frequently used in lieu of relative abundances or population estimates because they take less time and effort to generate, particularly for rare species (Kéry and Schmidt 2008). In rivers and streams, occurrences usually represent multiple individuals sampled from a watershed (or different flow paths within larger watersheds), such that a localized, catastrophic extirpation event at one occurrence would not necessarily affect the others. There is an increasing appreciation of the need to account for imperfect detection, which results in false absences in monitoring datasets (Kéry and Schmidt 2008; Haynes et al. 2013). Absence of a species when it is not observed from a given area is ensured only when the probability of observation is 1, a condition rarely satisfied in surveys, particularly with elusive targets such as fish (Peterson and Bayley 2004; Haynes et al. 2013). Detectability is a complex function of the probability of individual capture (which varies widely with sampling method and effort) fish size, physical habitat, local density, and seasonal or behavioral patterns (Bailey et al. 2004). Rarity itself increases the chances of missing present species 3 and imperfect detection can become inflated if a species is cryptic (Mackenzie et al. 2006), has particularly discontinuous or patchy distributions (Angermeier et al. 2002) or if habitats used by rare animals are difficult to sample efficiently (Parsley et al. 1989). The probability of detecting a species is therefore an intertwined combination of sampling gear performance and imperfect detection (Bailey et al. 2004) because aspects of both the sampling and target organisms affect eventual detection probabilities (Dolan and Miranda 2003). Imperfect detection can be accounted for in an occupancy-modeling framework using repeated site surveys to generate presence-absence histories for habitat patches (MacKenzie 2005). Presence-absence data has been used for studies of distribution and range (MacKenzie